DeepSeek/DeepGEMM
1
0
Fork 0
mirror of https://github.com/deepseek-ai/DeepGEMM synced 2025-05-06 07:14:22 +00:00
Code Issues Actions Packages Projects Releases Wiki Activity

Use c++20

Browse Source
This commit is contained in:
Chenggang Zhao 2025-04-03 15:47:59 +08:00
parent 6db7e1863b
commit 3a5539b7db
1 changed files with 1 additions and 1 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

2
deep_gemm/jit/compiler.py
View File

@ -96,7 +96,7 @@ def put(path, data, is_binary=False):
def build(name: str, arg_defs: tuple, code: str) -> Runtime:
# Compiler flags
nvcc_flags = ['-std=c++17', '-shared', '-O3', '--expt-relaxed-constexpr', '--expt-extended-lambda',
nvcc_flags = ['-std=c++20', '-shared', '-O3', '--expt-relaxed-constexpr', '--expt-extended-lambda',
'-gencode=arch=compute_90a,code=sm_90a',
'--ptxas-options=--register-usage-level=10' + (',--verbose' if 'DG_PTXAS_VERBOSE' in os.environ else ''),
# Suppress some unnecessary warnings, such as unused variables for certain `constexpr` branch cases